Rule-Based Information Extraction for Structured Data Acquisition using TextMarker

Information extraction is concerned with the location of specific items in (unstructured) textual documents, e.g., being applied for the acquisition of structured data. Then, the acquired data can be applied for mining methods requiring structured input data, in contrast to other text mining methods that utilize a bag-of-words approach. This paper presents a semi-automatic approach for structured data acquisition using a rule-based information extraction system. We propose a semi-automatic process model that includes the TEXTMARKER system for information extraction and data acquisition from textual documents. TEXTMARKER applies simple rules for extracting blocks from a given (semi-structured) document, which can be further analyzed using domain-specific rules. Thus, both low-level and higher-level information extraction is supported. We demonstrate the applicability and benefit of the approach with two case studies of two realworld applications.

• Type: Conference
• Year: 2008
• Where: LWA
• Authors: Martin Atzmüller, Peter Klügl, Frank Puppe
• Keywords: Information Extraction, LWA 2008, Software Engineering, Structured Data, Textual Documents